Forest Policy & Economics, June 2004, v. 6 iss. 3/4, p229-241. Available From:Link to Source
This paper explores how individual behaviour responds to externally imposed regulations or community self-governance in the context of extracting a common natural resource for personal economic gain (tragedy of the commons). The authors conduct a series of economic experiments in the field to compare how the changes in individual behaviour differ from game-theoretical predictions of the same. The authors first develop Nash equilibrium estimates of optimal individual extraction levels, to which the real-life results of the field experiments are compared.
A sample of 265 villagers was divided into groups of five individuals to participate in different experimental scenarios. Individuals’ payoffs were based on the extraction of a community resource, with the caveat that payoffs decrease as the aggregate extraction by the community increases. Each game participant was provided a payoff table revealing the expected payoff, given an individual player’s level of extraction and the group’s aggregate extraction. Individual extraction decisions were kept confidential, but the total group extraction amount was revealed after each round. After 10 rounds of the game, groups were provided a new set of rules for the following 10 rounds. Some groups were provided with a self-governing mechanism, allowing a brief group discussion prior to confidential individual extraction decisions in each round. Other groups were assigned an externally enforced penalty, of either US$50 or US$175, for over-extraction, and were informed of the social optimum extraction level of one unit. Finally, some groups were allowed to continue playing with no new rules, to serve as the control baseline.
According to game theory, the authors expected each individual to extract eight units absent any enforcement mechanisms, as well as in the case of nonbinding group discussions. The US$50 and US$175 penalties were expected to drop the individual extraction levels to six and one units, respectively.
Across the treatment groups, individuals extracted an average of four to five units throughout the first 10 rounds, less than the theoretical individual optimum of eight units. For the baseline control groups, this average rate of extraction was consistent across the entire 20 rounds of the game. Among the groups in which repeated group discussions were implemented, individual extraction levels decreased to a sustained average extraction level of two to three units. Finally, for the groups in which an external penalty was imposed, the reduction in individual extraction levels did not differ greatly depending on the amount of the penalty. Individuals in groups with the high-penalty treatment extracted 1.34 units on average, more than the 1 unit predicted by game theory; whereas individuals in groups with the low-penalty treatment extracted 3.32 units on average, less than the predicted 6 units.
Distributions of extraction data for each groups’ final 10 rounds revealed, however, that the social optimum of one unit was selected in over 40 percent of individual decisions in all but the baseline control group. This finding demonstrates that although the socially optimum level was not revealed to the community discussion groups, these groups were able to come to that realization through their self-enforcement discussions. It also suggests that for those groups on which an external penalty was imposed, knowledge of the optimum social extraction level was a greater driver of individual behaviour than the amount of the over-extraction penalty.
These results demonstrate the importance of taking into account the role of preferences in human behaviour when designing institutions, as well as the capacity of communities to effectively self-regulate. The policy implications are especially evident when considering the resources required to implement the monitoring and enforcement required by externally imposed regulations.